Bridging the Question-Answer Gap in Retrieval-Augmented Generation: Hypothetical Prompt Embeddings

• 类型：arxiv
• 标识：2607.29402
• 链接：http://arxiv.org/abs/2607.29402v1
• 主分类：rag
• 形态：method
• 被引：7
• 被引来源：Semantic Scholar
• S2被引：7
• 影响力被引：0
• TLDR：This work proposes Hypothetical Prompt Embeddings (HyPE), a framework that shifts the generation of hypothetical content from query time to the indexing phase, and transforms retrieval into a question-question matching task, bypassing the need for runtime synthetic answer generation.
• 待LLM分类：否
• 标题中文：弥合 RAG 中的问答差距：假设提示嵌入
• TLDR中文：提出Hypothetical Prompt Embeddings (HyPE)，将假设内容的生成从查询阶段前移到索引阶段，把检索转化为问题-问题匹配任务，无需运行时合成答案生成。
• 来源文件：
• /inbox/tom/_candidates/2026-08-03-agent-rag-longcontext-candidates.json
• /inbox/tom/_candidates/2026-08-04-agent-rag-longcontext-candidates.json
• [S2 enrich]